English

The SpeakIn System for VoxCeleb Speaker Recognition Challange 2021

Sound 2021-09-07 v1 Audio and Speech Processing

Abstract

This report describes our submission to the track 1 and track 2 of the VoxCeleb Speaker Recognition Challenge 2021 (VoxSRC 2021). Both track 1 and track 2 share the same speaker verification system, which only uses VoxCeleb2-dev as our training set. This report explores several parts, including data augmentation, network structures, domain-based large margin fine-tuning, and back-end refinement. Our system is a fusion of 9 models and achieves first place in these two tracks of VoxSRC 2021. The minDCF of our submission is 0.1034, and the corresponding EER is 1.8460%.

Keywords

Cite

@article{arxiv.2109.01989,
  title  = {The SpeakIn System for VoxCeleb Speaker Recognition Challange 2021},
  author = {Miao Zhao and Yufeng Ma and Min Liu and Minqiang Xu},
  journal= {arXiv preprint arXiv:2109.01989},
  year   = {2021}
}

Comments

Submitted to INTERSPEECH2021 VoxSRC2021 Workshop